Date

Attendees

Discussion items

WhoNotes
Ruslan Ganeev
  • Currently, all performance scripts are ready to be started. Control and monitoring are available from UI.
    • Longevity stand is ready, Locust and Graphana are configured to check the results.
    • There are 2 task sets: simple transactions and MST transactions.
    • We will have a meeting to share the current state tomorrow
  • Set up the system with shaper between two nodes. The current channel is 1 Kb/sec. Then Iroha start work in strange mode. 
    • After async GRPC we are making "finish" action. Even on a slow channel, everything works well
    • However, 1 block there is applying in 3 seconds. And in specific situation ordering gate can not get proposals. Probably it is related to packet loss.
    • Will try to continue testing with more significant packet loss. 
    • There is an open question – how the network will behave in case of 10 nodes, where 4 of them are with a bad connection.
    • Still looking for the source of that problem. There was one idea, but it was not approved.
  • My plan: understand the place when Iroha halts on 2 nodes, and then prepare the special build for Konrad to check on his network.

Action items

  • Vadim Reutskiy Prepare description of problem with backup and possible solutions